Download Subtitles for OCI Full Stack DR Oracle Kubernetes Video

hello everyone I'm surj rames product manager for OCA fullstack Dr joining me today is Anon M A fullstack OCA specialist from the emia cloud engineering team here at Oracle in this video we will Prov a quick overview of the oci OK integration with full stackd followed by a detailed demo OC kubernetes engine is a fully managed kubernetes service on OC for developing scaling and managing contain application with this new feature you can add the OK clusters as a member in the fullstack Dr protection group OK supports multiple note types such as manage notes and virtual noes fullstack DS support these four types once everything is available you can add the primary OK cluster to the primary Dr protection group and the standby OK cluster to the standby Dr production group based on the application Dr type you must include other resources like volume groups F storage Oracle database Etc in the respective Dr production groups finally create Dr plants in the standby Dr production let's look at some of the important features first periodically backup of all the cus resources including deployment State full sets and Crown jobs along with custom resources ensuring e recovery full stack Dr automatically replicate the container mystery images from one region to another region for seamless recovery it helps to enable automatic scale up and scale down of Lo pools in both primary and standby clusters based on the defined properties persons and volumes are included in Dr plans Ensure uring stateful applications can recover with data Integrity finally it provides a seamless integration with various OC resources such as load balancer networking article database for a comprehensive D strategy with this I will hand it over to Anton for a detailed demo of this feature to demonstrate the capabilities of this new feature I have set up a Nook cluster in duai as the primary deployment with an additional standby cluster located in Abu Dhabi in this setup we are running a stateless deployment on OK fully connected to an autonomous database the autonomous database is configured with autonomous Data Guard to ensure continuous data replication to the remote region full stackd will regularly take a backup of the application deployment and transfer the backup to the remote object storage pocket also as you can see in the diagram the work the number of worker nodes across both both clusters is not identical this will help on optimizing cost for the standby architecture I will show you how to set up full stack Dr to scale the worker nodes during the recovery process this is not mandatory but if you are in the case of reducing cost on the standby region this will be helpful for you as part of the switch over full stack Dr will first stop the application in the primary OK cluster then perform the switch over of the autonomous database it will scale the worker nodes accordingly and finally restore the backup from the bucket on into the OK cluster in the standby region and the backup will be rescheduled in the other way to Dubai region let's move now to the oci console demonstrate this feature in action you can see the cluster in Dubai region with one note pool and in the note pool I have uh four nodes and in Abu Dhabi the cluster in Abu Dhabi with the same note pool but with only two nodes and if I show you also the autonomous database in Dubai a primary DB in Dubai which is configured with autonomous Data Guard to replicate the data to Abu Dhabi region let's show now the uh deployment if you have the application already deployed in Dubai in Abu Dhabi there's no application running in in Abu Dhabi okay and this demo application is reachable from the internet with a public URL which is this one and uh it will be switched to the public load balancer in the Abu Dhabi region during the switch over let's move on and configure fullstack Dr in the primary region by creating the Dr protection group I will call it ok dxp I will put a bucket for the uh execution logs and I will create one I will do the same in the other region then I will do the association as usual so this one will be Associated will have the primary rule and the other one in Abu Dhabi assciation is done now and let's start by adding the oke into the primary Dr protection group so here you will select OK as resource type I will choose the primary OK I will choose the bucket for the backup and then I will schedule a daily backup I will choose this time and I will keep let's say for example three uh days for the retention okay I will not enable I don't have image replication I don't have any secret to replicate and I will choose the name space to be backup by default it's a full backup but I want to select only this the name space to be backup and and then I will choose the peer OK cluster where this backup will be restored then in the advanced option you can see the node pool scaling so here for this node pool I will put two as the minimum and four as the maximum okay if I have multiple note pool I can add multiple note pools and the other options I'm not using it but if you're using jump host or the native inv controller you can do the mapping here or any Vault mapping you can do them in this Advanced option now this is done I will hit add the is now added let's add now the primary autonomous database and I will choose a snapshot standby for the drill and I will hit that so the autonomous was added let's move to the standby region and add the members the RPG I will choose okay first I will select the standby OK I will select the bucket or the backup as it's a standby one I don't have to specify the uh backup schedule as of now I just have to map with the p okayy cluster and I willit and finally let's add the standby autonomous and again put snapshot stand by now for this deployment all the resources are now added to the both the protection group let's now create the Dr plan and as usual from the standby drpg let's start by the switch over plan I will give a name switch over and I will hit create so the switch plan is now created let's have a look so uh as you can see it's created with all the steps related to stopping and cleaning the the cluster in the primary region switch over the DB scale up the cluster in the standby restore the uh deployment in the standby and then reschedule the backup on the other way and scale down the old primary uh cluster this is the switch over let's create another plan the failover plan similarly the failover plan is now created and these steps are slightly different it's a failover of the autonomous scale up and restore because it's a real catastrophic event in case of fail and finally let's also create a uh drill type of plan the drill is also created it and as as usual it has the different steps let's focus now on the switch over for this demo uh let me run the pre-check I don't have any customization because I have DNS steering policies that will uh switch the traffic to the uh standby load balancers once I executed the switch over let me run the pre-checks and we can it's in the queue we can monitor from the logs here the pre-checks completed 3 minutes 4 uh 6 second let's now run the switch over ignore the pre-checks as was executed before and then execute the plan let indic progress so the part related to stopping the application on the primary OK cluster is completed it's on the stage of switching the autonomous DB if I check my application now here it's it should be unavailable okay and if I check the OK deployment in Dubai I will see that there's no more new shop application deployed okay so the plan is still in progress the we are at the phase of scaling up the Oki cluster in the standby as you can see here it was two nodes and it it's had been scaled up to four nodes now and the autonomous DB was switched over so it's also the application is now restored on the stand by Oki cluster which I should see all the pods here here we go and normally I should have my application back again from the uh standby region So the plan completed successfully everything is working fine and I can check again on primary that also the no more application is running on this OK e cluster and the number of nodes was also reduced to two okay where I have here in the PO noes with the application running here in Abu Dhabi so that's it for this video I hope you found it helpful and provided valuable Insight thank you for watching